Disease and damage at the level of Descemet's membrane/corneal endothelium can result in production of retrocorneal fibrous membrane (RCFM). Corneal endothelium in vivo responds to diverse types of pathology by converting to fibroblast-like cells. Experimentally induced RCFMs in rabbit eyes produce type I collagen as the predominant species, in contrast to the type IV collagen synthesized by corneal endothelial cells. Furthermore, in an in vitro model in which polymorphonuclear leukocytes (PMNs) modulate type IV collagen-synthesizing endothelial cells to type I collagen-synthesizing cells such modulated endothelial cells share phenotypic characteristics with the cells in retrocorneal fibrous membranes; the major collagen is type I, which can form fibrous interstitial extracellular matrices between multiple layers of the modulated cells. Thus, corneal endothelial modulation is involved in phenotypic switches in collagen gene expression. The control mechanism of type I collagen, the major modulated phenotype, has now been partially characterized as post-translational regulation: type I procollagen is synthesized in corneal endothelial cells; however, the molecule is not secreted into the extracellular matrix. This suggests that the new synthesized type I procollagen in this tissue must undergo intracellular degradation under physiologic conditions. The mechanism of selective intracellular degradation of type I procollagen in corneal endothelial cells will be studied with the known heat-shock protein (Hsp47), which is involved in type I procollagen synthesis, and ubiquitin, which is a major degradation pathway for selective intracellular protein. Simultaneously, we have also shown that bFGF is the direct mediator for corneal endothelial modulation while corneal endothelium modulation factor (CEMF) released by PMNs plays a role as an inducer of de novo synthesis of bFGF. Furthermore, we have demonstrated that the mere presence bFGF in Descement's membrane is sufficient neither for enhancing cell proliferation nor for modulating phenotypes of corneal endothelial cells. bFGF is now present in four isoforms which have differential subcellular locations and have been proposed to have differential biological activities. We therefore propose to study the molecular mechanism of bFGF in corneal endothelial modulation. Lastly, we propose to purify corneal endothelium modulation factor by using molecular cloning techniques. The long-term objective of this research is to understand the molecular mechanism of corneal endothelial modulation that inevitably leads to RCFM formation and to provide a basis for therapeutic intervention to prevent RCFM.